Reaction of BH4- with {Mo2Cp 2(μ-SMe)n} species to give tetrahydroborato, hydrido or dimetallaborane compounds: Control of product by ancillary ligands

Article abstract of DOI:10.1039/b407564a

The reaction of mono- or dichloro-dimolybdenum(m) complexes [Mo ₂Cp₂(μ-SMe)₂(μ-Cl)(μ-Y)] (Cp=η⁵-C₅H₅; 1, Y = SMe; 2, Y = PPh ₂; 3, Y = Cl) with NaBH₄, at room temperature gave in high yields tetrahydroborato (8), hydrido (9) or metallaborane (12) complexes depending on the ancillary ligands. The correct formulation of derivatives 8 and 12 has been unambigously determined by X-ray diffraction methods. That of the hydrido compound 9 has been established in solution by NMR analysis and confirmed by an X-ray study of the μ-azavinylidene derivative [Mo ₂Cp₂(μ-SMe)₂(μ-PPh₂)(μ-N= CHMe)] (10) obtained from the insertion of acetonitrile into the Mo-H bond of 9. Reaction of NaBH₄, with nitrile derivatives, [Mo₂Cp ₂(μ-SMe)_4-n(CH₃CN)_2n] ⁿ⁺ (5, n= 1; 6 =2), afforded the tetrahydroborato compound 8, together with a μ-azavinylidene species [Mo₂Cp₂(μ- SMe)₃(μ-N=CHMe)] (14), when n= 1, and the metallaborane complex 12, together with a mixed borohydrato-azavinylidene derivative [Mo ₂Cp₂(μ-SMe)₂(μ-BH₄)(μ-N= CHMe)] (13), when n=2. The molecular structures of these complexes have been confirmed by X-ray analysis. Preparations of some of the starting complexes (3 and 4) are also described, as are the molecular structures of the precursors [Mo₂Cp₂(μ-SMe)₂(μ-X)-(μ-Y)] (1, X/Y = Cl/SMe; 2, X/Y = Cl/PPh₂; 4, X/Y = SMe/PPh₂).

Full text of DOI:10.1039/b407564a

View Article Online / Journal Homepage / Table of Contents for this issue
F U L L P A P E R
Reaction of BH4− with {Mo2Cp2(-SMe)n} species to give
tetrahydroborato, hydrido or dimetallaborane compounds:
control of product by ancillary ligands
Nolwenn Cabon,a François Y. Pétillon,*a Philippe Schollhammer,*a Jean Talarmina and
Kenneth W. Muir*b
a
UMR CNRS 6521, Chimie, Electrochimie Moléculaires et Chimie Analytique,
UFR Sciences et Techniques, Université de Bretagne Occidentale, CS 93837,
29238 Brest-cedex 3, France. E-mail: francois.petillon@univ-brest.fr; schollha@univ-brest.fr
Chemistry Department, University of Glasgow, Glasgow, UK G12 8QQ.
b
E-mail: ken@chem.gla.ac.uk
Received 19th May 2004, Accepted 30th June 2004
First published as an Advance Article on the web 23rd July 2004
The reaction of mono- or dichloro-dimolybdenum(III) complexes [Mo2Cp2(-SMe)2(-Cl)(-Y)] (Cp=5-C5H5; 1,
Y = SMe; 2, Y = PPh2; 3, Y = Cl) with NaBH4 at room temperature gave in high yields tetrahydroborato (8), hydrido (9)
or metallaborane (12) complexes depending on the ancillary ligands. The correct formulation of derivatives 8 and 12
has been unambigously determined by X-ray diffraction methods. That of the hydrido compound 9 has been established
in solution by NMR analysis and confirmed by an X-ray study of the -azavinylidene derivative [Mo2Cp2(-SMe)2(-
PPh2)(-NCHMe)] (10) obtained from the insertion of acetonitrile into the Mo–H bond of 9. Reaction of NaBH4 with
nitrile derivatives, [Mo2Cp2(-SMe)4−n(CH3CN)2n]n+ (5, n=1; 6 n=2), afforded the tetrahydroborato compound 8, together
with a -azavinylidene species [Mo2Cp2(-SMe)3(-NCHMe)] (14), when n=1, and the metallaborane complex 12,
together with a mixed borohydrato-azavinylidene derivative [Mo2Cp2(-SMe)2(-BH4)(-NCHMe)] (13), when n=2.
The molecular structures of these complexes have been confirmed by X-ray analysis. Preparations of some of the starting
complexes (3 and 4) are also described, as are the molecular structures of the precursors [Mo2Cp2(-SMe)2(-X)-
(-Y)] (1, X/Y = Cl/SMe; 2, X/Y = Cl/PPh2; 4, X/Y = SMe/PPh2).
derivatives of the dimolybdenum starting complexes. Some of the
1. Introduction
results demonstrate an understandable competition between borane
or H2 elimination and dimolybdenum polyborohydride dispropor-
tionation, giving rise to molybdenum hydrides or molybdoboranes.
Furthermore, since the ancillary ligands condition the reactivity
The thermal stability of covalent f-element borohydrides
allows them to be isolated and characterised easily.1 In contrast,
borohydride complexes [M(BH4)] of d-transition metals, with
the exception of those of group 4, are less stable, decomposing
readily to give either the corresponding hydride after elimination of
borane2,3 or a metallaborane after loss of dihydrogen.3,4 Neverthe-
less, several stable borohydride complexes of d-transition elements
have been isolated over the last fifteen years: examples are now
known for the metals niobium,5 chromium,6 molybdenum,7 manga-
nese,8 iron,9 ruthenium,10 osmium,10b rhodium11 and iridium.12
The usual synthetic route to a borohydride complex involves
reaction of a metal halide with an alkali-metal borohydride.13 Less
common routes to borohydride compounds are based on either
(i) insertion of a hydridoborane into a metal-hydride bond14 or
(ii) reaction of a borate salt with a cationic transition metal com-
plex containing labile dative ligands in what is formally a substitu-
tion process.7,10c However, when an alkali-metal borohydride reacts
with a metal halide which contains more than one chloro ligand, a
metallaborane can be obtained via a stepwise process: initial forma-
tion of a transient polyborohydride is followed by elimination of
dihydrogen.15 It is also well established that replacement of Cl− by
−
of the metal reagent toward BH4 , we have determined, where it
is appropriate, the molecular structures of the quadruply-bridged
dimolybdenum precursors. Part of this work has been published in
preliminary form.7
2. Results and discussion
2.1. Synthesis and characterisation of precursors 1–7.
Molecular structures of 1, 2 and 4
The dinuclear complexes selected for reaction with sodium
borohydride fall into two groups (see Fig. 1). Complexes
1–4 make up a series of quadruply-bridged bimetallic complexes
[Mo2Cp2(-SMe)2(-X)(-Y)] in which the two halide or pseudo-
halide bridging ligands, X and Y, have been systematically
varied in order to assess their influence on the reactivity of the
−
complexes towards BH4 ; the respective X/Y combinations are
Cl/SMe, Cl/PPh2, Cl/Cl, and SMe/PPh2. The second group consists
of the thiolato-complexes 5 and 6 in which chloride ligands are
absent but two or four labile nitrile ligands are present. These
complexes were chosen in order to confirm that substrates which
do not contain halide complexes can react with BH4− to give either
borohydride compounds or metallaboranes. The tetrakis(nitrile)
complex [Mo2Cp2(-SMe)2(MeCN)4](BF4)2 (5) and the bis(nitrile)
derivative [Mo2Cp2(-SMe)3(MeCN)2] (BF4) (6) have previously
been synthesised and characterised by spectroscopy and elemental
analysis; for (5) the structure has also been confirmed by X-ray
diffraction.16,17
−
the pseudohalide BH4 gives either a stable hydroborato complex
or a metal hydride if the initial substitution reaction is followed by
loss of BH3.2
At present it is impossible to state reliably which factors favour
the formation of a borohydride rather than a hydrido complex in
such reactions. We decided to approach this problem by reacting
borate salts with bimetallic reagents in which the bridging units
are systematically varied. Accordingly, we describe here the
reactions of sodium borohydride with both the bimetallic halide
and pseudohalide complexes [Mo2Cp2(-SMe)2(-X)(-Y)]
(Cp = 5-C5H5; X/Y = Cl/SMe, Cl/PPh2, Cl/Cl, and SMe/PPh2)
and the related cations [Mo2Cp2(-SMe)4−n(CH3CN)2n]n+ (n = 1 or
2) which contain labile nitrile ligands. One successful outcome of
these reactions is the synthesis of borohydride and metallaborane
2.1.1. Molecular structures of 1 and 2. We earlier reported
the synthesis of the hetero-quadruply-bridged complex [Mo2Cp2(-
SMe)3(-Cl)] (1) and its characterisation by spectroscopy and
2 7 0 8
D a l t o n T r a n s . , 2 0 0 4 , 2 7 0 8 – 2 7 1 9
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4

View Article Online
Fig. 2 A view of a molecule of 1. Here and elsewhere 20% probability
ellipsoids are shown. Selected angles (°): Mo21–S21–Mo22 64.1(1),
Mo21–S22–Mo22 64.0(1), Mo21–S23–Mo22 63.5(1), Mo21–Cl21–Mo22
63.2(1).
Fig. 1 Dimolybdenum precursors relevant to this work.
elemental analysis.18 However, in the absence of suitable crystals,
its molecular structure could not by confirmed by X-ray diffraction.
We have also described the preparation of the quadruply-bridged
compound [Mo2Cp2(-SMe)2(-Cl)(-PPh2)] (2), which is obtained
quantitatively by the addition of excess diphenylphosphine HPPH2
to a tetrahydrofuran solution of 1. 2 was also characterised by
spectroscopy and elemental analysis but not by X-ray diffraction.19
Several attempts to get crystals of 2 of diffraction quality from
a preparative sample of the complex in diethylether solution
at −15 °C failed. Finally, a crystal of poor quality was obtained; it
turned out to have an asymmetric unit which contained a molecule
of 2 co-crystallized with a molecule of the parent species 1. This
result was unexpected, as the parent compound 1 was shown by
NMR techniques to be only a minor impurity in the solution from
which the crystal was obtained. The presence of molecules of
[Mo2Cp2(-SMe)3(-Cl)] (1) and [Mo2Cp2(-SMe)2(-Cl)(-PPh2)]
(2) in the same crystal has enabled us to complete a series of X-ray
diffraction studies of quadruply-bridged dimolybdenum(III) deriva-
tives with a variable number of thiolato ligands.
Fig. 3 A view of a molecule of 2. Selected angles (in °): Mo11–S11–
Mo12 64.4(1), Mo11–S12–Mo12 64.5(1), Mo11–Cl11–Mo12 62.4(1),
Mo11–P11–Mo12 66.5(1).
Molecules of 1 and 2 (Figs. 2 and 3, Table 1) contain pairs of
molybdenum atoms, in oxidation state III, directly bonded to each
other and symmetrically bridged by four anionic ligands. In 1 the
two molybdenum atoms are bridged by one chloro and three SMe
ligands. Replacement by PPh2 of the SMe trans to Cl converts 1
into 2. If the Mo–Mo bonds are ignored, the coordination of the
metal atoms can be described in terms of a four-legged piano
stool. Though many quadruply-bridged Mo(III)–Mo(III) com-
plexes have been characterised by crystallography, in most cases
the bridging ligands are either sulfur20 or halogen21–23 donors; of
these only one, [Mo2Cp′(-Cl)3(-PPh2)] (Cp′ = 5-C5Me4Et),
also contains a bridging phosphido group.21b Complexes 1 and 2
therefore provide the first structural examples of quadruply-bridged
dimolybdenum(III) thiolate derivatives in which there are bridging
chloro and/or phosphido ligands. However, the molecular structure
of a mixed thiolato–phosphido–sulfido–tetrabridged compound
[Mo2Cp(-SiPr)(-S)2(-PPh2)] has been described, but in this
complex the Mo atoms are in oxidation state IV.24
or pseudohalide (X) ligands (X = SR,20 Cl,21 Br,22 I,23 Cl/PPh2,21b
Br/SR26 andI/SR26).TheseMo(III)–Mo(III)distancesappearsensitive
to the nature of the bridging ligands: shorter distances are found for
tetrathiolato- and tetrachloro-bridged derivatives where the mean
values are respectively 2.595 and 2.601 Å;20,21 the longest bonds
occur in tetraiodo-bridged compounds [2.708(9) Å];23 intermediate
distances are present in tetrabromo- and mixed chlorophosphido-
and tris(thiolato)bromo- or iodo-bridged complexes for which the
respective means are 2.643(2),22 2.638(1)21b and 2.622(1) Å.26 The
Mo–S bonds distances (average 2.461 Å) and Mo–S–Mo angles
(average 64.1°) in 1 and 2 are in the ranges (2.45–2.46 Å and
63–64°) generally observed for the tetra(-thiolato) compounds.20
Evidently, replacement of one or more thiolates by a pseudohalide,
such as chloride or phosphide, has little influence on these para-
meters. However, the most striking comparison is that between
[Mo2Cp2(-SMe)2(-Cl)(-PPh2)] (2) and the analogous compound
[Mo2Cp2′(-Cl)3(-PPh2)] (Cp′ = C5Me4Et).21b These compounds
have the same trans arrangement of Cl and PPh2 ligands and
differ only in the nature of the bridging ligands [X = Cl or SMe]
cis to PPh2 and in the substituents of the cyclopentadienyl groups.
It is noteworthy that the Mo–Mo distances [2.627(1) Å in 2 and
2.639(1) Å in the Cp′ complex] are nearly identical and that the
mean Mo–Cl distance of 2.510(2) Å in the Cp′ compound is only
slightly [0.026(4) Å] shorter than the corresponding mean for 2,
For the two diamagnetic neutral species 1 and 2 a direct
metal-metal bond is in accord with the short Mo–Mo distances
of 2.601(1) and 2.627(1) Å, respectively, but is also compatible
with acute Mo–Y–Mo (Y = S, Cl, P) bridging angles in the range
62.4(1)–66.5(1)°.25 The metal-metal distances in 1 and 2 are in the
range [2.580(2)–2.708(3) Å] found for related quadruply-bridged
Mo(III)–Mo(III) compounds containing three-electron-donor halide
D a l t o n T r a n s . , 2 0 0 4 , 2 7 0 8 – 2 7 1 9
2 7 0 9

View Article Online
Table 1 Selected distances (in Å) and angles (in °) in [Mo2Cp2(-SMe)2(-X)(-Y)] complexesa
1
2
4
10b
12ac
13
X
Y
Cl
Cl
SMe
PPh2
NCHMe
PPh2
B2H6
NCHMe
SMe
PPh2
BH4
Mo–Mo
Mo–S
Mo–S
Mo–S
Mo–PPh2
Mo–Cl
Mo–N
2.601(1)
2.452(3)
2.465 (3)
2.449(3)
2.627(1)
2.454(3)
2.454(3)
2.647(1)
2.492(1)
2.462(1)
2.469(1)
2.415(1)
2.612(1)
2.471(1)
2.470(1)
2.717(1)
2.428(1)
2.435(1)
2.630(1)
2.469(1)
2.444(1)
2.458(3)
2.476(3)
2.455(3)
2.475(3)
2.471(3)
2.472(1)
2.469(1)
2.477(1)
2.439(1)
2.469(1)
2.436(1)
2.398(3)
2.548(3)
2.394(3)
2.524(3)
2.414(1)
2.116(3)
2.481(3)
2.486(3)
2.083(2)
2.094(2)
a In 1, 2, 4 and 10 S and S are mutually trans, as are X and Y. In the bis(-thiolates) 12a and 13 S and S are mutually cis. b The molecule has exact C2
symmetry. c The B2H6 ligand occupies both X and Y sites in 12a and the molecule straddles a crystallographic mirror plane.
whereas the mean Mo–P distance dramatically increases from
2.396(2) Å in 2 to 2.458(2) Å in the Cp′ compound. Both the slight
shortening of the Mo–Cl distances and the lengthening of the Mo–P
distances probably result from the combined electronic effects of
the cyclopentadienyl groups and the equatorial halide or pseudo-
halide ligands.
Lastly, replacement of an SMe ligand in 1 by the sterically more
demanding PPh2 group in 2 has virtually no effect on the Mo–S(cis
to Cl) distances [respective means 2.462(3) and 2.463(3) Å]. This
contrasts with the lengthening of the mean Mo–Cl distance by
0.053(4) on going from 1 to 2 which suggests that the phosphido
ligand exerts a stronger trans-influence than SMe.
The molecule of 4 (see Fig. 4 and Table 1) contains two CpMo
fragments bridged almost symmetrically by one PPh2 and three
SMe ligands, so that each Mo atom has a four-legged piano-stool
coordination, as in 1 and 2. The Mo–Mo bond [2.647(1) Å] is a little
longer than those in 1 and 2, as are corresponding Mo–S distances.
A higher trans-influence for PPh2− relative to Cl− is consistent with
the Mo–S distances trans to these ligands in 1 and 4 (see Table 1).
Other structural parameters of 4 are unremarkable and agree with
those in comparable structures, in particular 1, 2 and [Mo2Cp′(-
Cl)3(-PPh2)](Cp′ = 5-C5Me4Et).21b The spectroscopic (1H and
31P{1H}NMR) data for 4 (see Experimental section) show that the
solution structure is consistent with that observed in the solid state.
In particular, the 31P NMR chemical shift of 4 ( 88.6 ppm/H3PO4)
confirms the presence of a metal-metal bond.27 The 1H NMR spec-
trum is typical for a complex with a {Mo2Cp2(-SMe)3} core: it
exhibits a single resonance for the two cyclopentadienyl ligands
and three peaks for the three SMe bridges.25 Moreover, the presence
of two phenyl groups in 4 is shown by characteristic multiplets at
7.3–7.0 ppm.
2.1.2. Synthesis and characterisation of 3. A solution of the
salt [Mo2Cp2(-SMe)2(CH3CN)4](BF4)2 (5)17 in dichloromethane
was treated with excess Et4NCl to afford the neutral bis(-chloro)-
bimetallic complex 3 as a green solid in approximately 71% yield.
(1)
Compound 3 has been characterised from spectroscopic and
1
analytical data (see Experimental section). It was shown by H
NMR spectroscopy that in solution the complex is present in two
isomeric forms, 3a and 3b, which differ only in the orientations
(syn and anti) of the bridging SMe groups. These isomers were
inseparable by conventional chromatographic techniques and were
obtained in 1:1 ratio.
2.1.3. Synthesis and characterisation of 4 and 7. Molecular
structure of 4. Attempts to crystallize the hydride compound
9 (see section 2.2 below) in diethylether at 8 °C under nitrogen
gave a mixture of crystals of two complexes 4 and 7 in about a
1:1 ratio according to a spectroscopic analysis (eqn. (2)). A red-
orange crystal of 4 was picked from the mixture and its structure
determined by X-ray analysis.
Fig. 4 A view of a molecule of 4. Selected angles (in °): Mo1–S1–Mo2
64.5(1), Mo1–S3–Mo2 65.0(1), Mo1–S2–Mo2 64.7(1), Mo1–P1–Mo2
66.1(1).
The formulation of compound 7 can be tentatively deduced from
the spectroscopic data (see Experimental section). The 31P{1H} spec-
trum of 7 exhibits just one signal in the region expected for bridg-
ing phosphido ligands ( 88.7 ppm/H3PO4).28 The observed shift is
typical for a phosphido group bridging two directly-bonded metal
atoms.27 The 1H NMR spectrum of 7 exhibits two different sets of
signals for the phenyl groups: an AB pattern at 8.23 ppm is present
in the region usual for aryl groups and eight-proton multiplets are
observed between 7.0 and 7.3 ppm. These data suggest strongly that
one phenyl group is bonded to a Mo atom through a carbon–carbon
(2)
2 7 1 0
D a l t o n T r a n s . , 2 0 0 4 , 2 7 0 8 – 2 7 1 9

View Article Online
double bond so that the environment at each Mo atom is different,
thereby explaining the non-equivalence of the cyclopentadienyl
ligands (two resonances at 5.57 and 5.42 ppm). The presence of one
signal (3 1H) in the range expected for a SMe group ( 2.82 ppm) is
also consistent with the structure proposed for 7.
nitrile in tetrahydrofuran under mild conditions gave quantitatively
the targeted -azavinylidene product 10; similarly, 9 reacted with
t-BuNC in tetrahydrofuran at room temperature, affording the
-formimidoyl derivative 11 in good yield (66%).
2.2. Reactions of dimolybdenum complexes with NaBH4
2.2.1. Reaction of 1 with NaBH4. [Mo2Cp2(-SMe)3(-
Cl)] (1) reacted with a large excess of sodium borohydride in
tetrahydrofuran at room temperature to afford in high yield (88%)
the borohydride compound 8 as a moderately air-sensitive orange
solid.
(5)
(3)
The structure of compound 10 was determined by X-ray analy-
sis (Fig. 5). The complex is structurally related to the -phosphido
species 2 and 4 (see Table 1). It consists of two CpMo fragments
which are directly connected by a Mo–Mo bond of 2.612(1) Å
in addition to being bridged by two thiolates, one phosphide and
the nitrogen atom of the azavinylidene ligand. A crystallographic
dyad axis passes through the atoms P1, N1 and C2, so that the
PPh2 and NC(H)Me ligands are disordered. All the bridges are
symmetrical, even when not constrained to be so, as is the case for
the single SMe ligand. The azavinylidene N1–C2 bond distance
[1.234(7) Å] accords with a bond order of two.31 The closest
structural comparison to 10 is provided by the quadruply-bridged
Mo(III)–Mo(III) compound [Mo2Cp2(-SMe)3(-NC(H)Me)] (14),7
which differs from 10 by having a -thiolato instead of a -phos-
phido ligand trans to the azavinylidene group. The mean Mo–N
distances in 14 and 10 [2.078(8) and 2.116(3) Å] demonstrate yet
again that phosphide exerts a high trans-influence (see sections
The characterisation of 8 was accomplished by NMR spectro-
scopy, which showed that it was identical to the tetrahydroboride
[Mo2Cp2(-SMe)3(-BH4)] we had previously obtained in
moderate yields, together with an azavinylidene product, from
the reaction of [Mo2Cp2(-SMe)3(CH3CN)2](BF4) with NaBH4 in
either tetrahydrofuran or acetonitrile.7 The synthesis of 8 from the
metal chloride 1 is significantly more convenient than our previous
method, which is based on the lability of CH3CN ligands.
2.2.2. Reaction of 2 with NaBH4 and of 9 with CH3CN and
t-BuNC. Since 2, like 1, is a -chloro-complex, it also seemed a
suitable candidate for reaction with BH4− to give the corresponding
-BH4 derivative. However, though the reaction of 2 with NaBH4
proceeded with good selectivity, the targeted tetrahydroboride
derivative was not obtained. Instead, reaction of excess NaBH4
with 2 in tetrahydrofuran at room temperature led first to a violet
solution; this rapidly turned green and from it there was extracted
in high yield (74%) after evaporation of solvent the hydrido com-
pound [Mo2Cp2(-SMe)2(-H)(-PPh2)] (9) as a very air-sensitive
green solid.
1
2.1.1 and 2.1.3 above). H NMR data for 10 are in accord with
the formulation proposed from the X-ray study. In particular, the
proton attached to the C2 atom of the azavinylidene ligand appears
at low-field ( 8.08) as a quartet (JH–H = 4.6 Hz) by coupling with
the methyl group also attached to C2. Formation of 10 involves
insertion of NC–CH3 into the Mo–H bonds of 9, followed by a
shift of the hydride onto the nitrile carbon atom to form a three-
electron-donor -NC(H)Me azavinylidene ligand. Similar
reactions have been reported previously.32 Finally, the molecular
structure of 10 supports that proposed from spectroscopic data for
its parent complex 9.
(4)
Additional confirmation of the structure of 9 comes from for-
mation of the -formimidoyl derivative 11 when 9 reacts with
isonitrile (eqn. (5)). This requires insertion of isonitrile into an
Mo–H bond, behaviour previously demonstrated by similar hydrido
derivatives.33 Although we could not get crystals of 11 which were
suitable for a diffraction study, the characterisation of 11 can be
carried out on the basis of the analytical, IR and NMR data. The IR
spectrum shows a (CN) band (1651 cm−1) in the range typical of
CN double bonds, viz., 1470–1690 cm−1.34 The 1H NMR spectrum
exhibits the sets of resonances expected for the {Mo2Cp2(-SMe)2}
core and for the t-Bu and Ph groups. The presence of two reso-
nances at 5.30 and 5.24 ppm assigned to the two cyclopentadienyl
ligands accords with the unsymmetrical geometry proposed for 11.
Moreover, a doublet ( JP–H = 4.4 Hz) detected at low field ( 11.18)
is typical of a formimidoyl proton.33 The formimidoyl bridge -
HCN(t-Bu) is also characterised in the 13C{1H} NMR spectrum
by the observation of one resonance ( 232.5), at low field in the
carbene range. Finally, the 31P shift (92.6 ppm) is in accordance with
the maintenance of the phosphido group in a bridging position28 on
transforming 9 to 11.
Compound 9 is also somewhat unstable in organic solvents; it is
transformed in a few hours into complexes 4 and 7, as mentioned
above. However, in the solid state its decomposition took a few
days. Nevertheless, 9 can be confidently characterised by elemental
analysis and 1H and 31P NMR spectroscopy. Thus, its 31P{1H} NMR
spectrum shows a single resonance at 154.3 ppm corresponding to
the -phosphido ligand.28 Moreover, the low-field chemical shift of
the phosphorus atom is consistent with the presence of a Mo–Mo
bond.27 Comparing this chemical shift with that of the parent com-
plex 2 ( 120.3),19 a valuable deshielding effect (ca 35 ppm) is
observed on the phosphido resonance when chloride is replaced by
hydride. The 1H NMR spectrum of 9 displays only one cyclopenta-
dienyl resonance ( 5.32), typical for a symmetrical molecule and
thus suggesting a bridging position for the hydride ligand. This is
confirmed by the high-field chemical shift ( −8.9) of the resonance
due to the hydrido ligand.29
Since we were unable to get suitable crystals of 9 for X-ray analy-
sis (see section 2.1.3), we decided to react this hydrido compound
with nitrile and isonitrile in order to confirm its proposed structure.
It is well known that the interaction of nitriles or isonitriles with
bimetallic hydrido compounds can lead to partial hydrogenation of
these ligands, with the respective formation of -alkylidenimido
and -formimidoyl ligands.30 As expected, reaction of 9 with aceto-
2.2.3. Reactions of 3 and 4 with NaBH4. The reaction of the
bis-chloro compound 3, [Mo2Cp2(-SMe)2(-Cl)2], with excess
NaBH4 at room temperature resulted in the formation of a single
metallaborane 12, which was isolated in high yield.
D a l t o n T r a n s . , 2 0 0 4 , 2 7 0 8 – 2 7 1 9
2 7 1 1

View Article Online
Table 2 Selected bond lengths (in Å) and angles (in °) for [Mo2Cp42(-
SMe)2(-B2H6)] (12a)
Mo1–Mo2
Mo1–S1
Mo1–B1
Mo1–H2
B1–H1
2.717(1)
2.428(1)
2.406(3)
1.88(3)
1.17(3)
1.10(3)
Mo2–S1
Mo2–B1
Mo2–H1
B1–H2
2.435(1)
2.402(3)
1.90(3)
1.17(3)
1.700(7)
B1–H3
B1–B1a
Mo1–S1–Mo2
H1–B1–H3
B1a–B1–H1
B1a–B1–H3
69.9(1)
103(2)
117(1)
118(2)
H1–B1–H2
H2–B1–H3
B1a–B1–H2
101(2)
103(2)
114(2)
a x, −y, z.
MoHB bridges. The length of a direct MoIII–B bond was found to
be 2.21 Å in [Cp*2Mo2(Cl)2(-B3H7)].4c The B–B bond distance in
12a is close to the corresponding distances in the related complexes
[Ta2(p-tolNCHNtol-p)4(-B2H6)] [1.68(2) Å]35 and [Mo2Cp2*(-
Cl)2(-B2F6)] (Cp* = 5-C5Me5) [1.63(3) Å],4c but is significantly
shorter than that found in [Ta2Cp2*(-Br)2(-B2H6)] [1.88(3) Å].4a
These results indicate that 12a is stabilised by covalent interac-
Fig. 5 A view of a molecule of 10. Atoms P1, N1 and C2 lie on a two-
fold axis. Disorder of the phosphido phenyl rings and of the azavinylidene
ligand is not show. Selected distances and angles (in Å and °): N1–C2
1.234(7), C2–C3 1.484(9), Mo1–S1–Mo1i 63.8(1), Mo1–P1–Mo1i 65.5(1),
Mo(1)–N(1)–Mo(1)i 76.2(1), Mo(1)–N(1)–C(2) 141.9(1). Symmetry code:
(i) −x, y, −z+1/2.
2−
tion of the {Mo2Cp2(-SMe)2}2+ moiety with the B2H6 anion
through four bent Mo–H–B bonds, only two being independent
[Mo–H 1.88(3) and 1.90(3); B–H 1.17(3) Å (both)]; the bridging
B–H bonds are somewhat longer than the terminal B–H bonds
[1.10(3) Å]. 12a can also be viewed as a metallaborane cluster since
it has the expected electron count (40 cve) for a M2B2 tetrahedron,
with two thiolates bridging the Mo–Mo bond and four hydrogen
atoms connecting the two boron atoms to the two molybdenum
atoms. Three related -hexahydrodiborato complexes have been
structurally characterised: [Ta2(p-tolNCHNtol-p)4(-B2H6)],35
[Ta2Cp2*(-Br)2(-B2H6)]4a and [Mo2Cp2*(-Cl)2(-B2H6)].4c The
structure of 12 differs significantly from the latter two of these
complexes which have direct metal–boron bonds. Moreover, in the
MoCp* complex an unsymmetrical arrangement of the B2H6 group
is observed.
(6)
Thenewcompoundwasidentifiedasamolybdenumdimerbridged
by an ethane-like B2H6 ligand on the basis of its spectroscopic data
and solid-state structure. It was shown by NMR spectroscopy that
12 was obtained as a mixture of two isomers which differ only in
the orientations syn (12a) or anti (12b) of the bridging SMe groups,
as discussed below. These isomers could not be separated by con-
ventional chromatographic techniques and were obtained in a 4:1
ratio. However, only the syn-isomer 12a was formed in a reaction
described in the next section, and crystals of 12a so obtained were
used for spectroscopic and X-ray studies of 12.
The 11B NMR spectrum of 12a contains only one resonance,
suggesting a single boron environment, and the chemical shift
( −44.1) accords with hydrogen-bridged rather than direct Mo–B
1
linkages. The H NMR spectrum shows a single Cp resonance
(relative intensity 10) at 4.47 ppm, one SMe signal (intensity 6) at
1.84 ppm, one B–H resonance (intensity 2) at 1.11 ppm, and a single
Mo–H–B signal (intensity 4) at −12.40 ppm. All these NMR data
are consistent with both a syn orientation of the SMe groups and
an ethane-like B2H62− ligand which symmetrically bridges the two
{CpMo} cores, as shown in eqn. (6). As formulated, 12 is derived
from [M2Cp2*(-Cl)2(-B2H6)] (M = Mo, W; Cp* = 5-C5Me5),4c,4d
with SMe− and Cp respectively replacing Cl− and Cp*. 12 also
resembles [Ta2Cp*2(-X)2(-B2H6)] (X = Cl, Br); however, the
tantalum complexes exist in solution as a mixture of isomers: the
predominant component in solution and in the solid contains an un-
symmetrical B2H6 ligand but at least one of the minor components
in solution may have a symmetrical B2H6.4a
This formulation of 12 as a -hexahydrodiborato complex was
confirmed by X-ray analysis of a single crystal of 12a, obtained
from a cooled diethylether solution (see Fig. 6 and Tables 1 and
2). The structure contains a fairly long Mo–Mo bond [2.717(1) Å]
which is bridged both by two thiolates in a syn orientation and by a
symmetrical B2H6 group with a B(1)–B(1)i separation [1.700(7) Å]
consistent with boron–boron bonding. A crystallographic mirror
plane contains the metal atoms and bisects the B–B bond. The
MoB distances [2.406(3) and 2.402(3) Å] are equal to within
experimental error and indicate the presence of four equivalent
Fig. 6 A view of a molecule of 12a.
As pointed out earlier, when 12 is formed by reaction of 3 with
1
NaBH4 (eqn. (6)), the H NMR spectrum indicates the presence
in solution of two inseparable isomers 12a and 12b. Since a syn
orientation of the SMe groups has been demonstrated for 12a, two
geometries are mainly possible for 12b (see Scheme 1): (i) an anti-
isomer with the hexahydrodiborato ligand symmetrically-bridged
(A), and (ii) a syn- or anti-isomer with the hexahydrodiborato ligand
unsymmetrically-bridged (B).
The H NMR spectrum of a pure sample of 12a and that of a
mixture of 12a and 12b are unchanged over the temperature range
223–293 K. This indicates that interconversion or fluxional pro-
cesses do not occur in solutions of 12. Subtracting the H NMR
1
1
spectrum of 12a from that of 12 yields the spectral pattern of 12b.
It shows only one resonance in the region expected for Cp ( 4.95;
2 7 1 2
D a l t o n T r a n s . , 2 0 0 4 , 2 7 0 8 – 2 7 1 9

View Article Online
of 8 with bases. Abstraction of BH3 from 8 to give the hydrido
compound [Mo2Cp2(-SMe)3(-H)] proved to be very difficult: the
usual BH3 abstractors failed to transform 8 into the corresponding
hydride. Thus, no reaction was observed when solutions of 8 and
triethylamine in either tetrahydrofuran or dichloromethane were
stirred overnight at room temperature; unchanged 8 was recovered
quantitatively (eqn. (9)). Heating a dichloromethane solution of 8
and Et3N to 60 °C only achieved the thermal decomposition of 8;
after 24 h the borohydride was completely transformed into uniden-
tified decay-products and no hydrido derivative was detected in the
NMR spectrum. Similar reactions of the borohydride complex 8
with Lewis bases, such as acetonitrile and isonitrile (t-BuNC), also
failed to transform 8 into the corresponding hydride.
Scheme 1 Possible geometries for 12b.
relative intensity 10), suggesting that the Cp groups are equivalent
in solution. The remaining pattern reveals two signals at 1.94 and
1.42 ppm (both of intensity 3) assigned to two SMe groups, and also
two kinds of MoHB bridges ( −12.02 and −15.45; both; intensity
2). Signals for terminal BH are not detected. Another interesting
feature is that the 11B NMR spectrum of 12b indicates the presence
of a single boron environment, and the chemical shift ( −44.9) is
more consistent with hydrogen-bridged rather than direct Mo–B
bonding. These observations eliminate structure (B) in which the
B2H62− ligand bridges the CpMo–MoCp core unsymmetrically. On
the other hand, (A) in Scheme 1 has two types of MoHB protons,
two types of SMe hydrogens, equivalent Cp groups, and equivalent
borons: all these features are consistent with the 1H and 11B NMR
data, since no atom exchanges have been detected on the NMR time
scale from 223 to 293 K.
(8)
Complex 12 has been obtained here in high yield and as a single
product from the reaction of the dichloro compound 3 with NaBH4
under mild conditions. In contrast, with a single exception,4e
previously reported syntheses of dimetallaboranes containing an
ethane-like B2H6 ligand have required harsh conditions or given
low-to-moderate yields.4a–4d,5a,35,36 Another interesting feature is that
12 is the first stable example of an 5-C5H5-metallaborane with a
B2H6 ligand; in all other known derivatives the cyclopentadienyl
ligands has at least one non-hydrogen substituent.4a,4c–4e,5a It is well
known that early transition metal monocyclopentadienyl halides
react readily with nucleophiles or Lewis bases.37 The high reactivity
(9)
As shown in eqn. (10), [Mo2Cp2(-SMe)2(CH3CN)4](BF4)2 (5)
reacted with 3 equivalents of NaBH4 at room temperature in aceto-
nitrile to give, after complete consumption of the starting material
(about 4 h), the mixed -borohydride--azavinylidene derivative
13 as the major product. 13 (yield: 30%) was separated by column
chromatography from a minor product, the dimetallaborane species
12a (yield: 10%), and several other compounds. These were
obtained in small amounts and could not be characterised.
−
of BH4 towards the dichlorodimolybdenum(III) compound 3 to
give the metallaborane derivative 12 is consistent with this obser-
vation. Moreover, [Cp*2M2(-Cl)4] species (M = MoIII, WIII) react
readily with electrophilic reagents such as BH3.thf to afford ethane-
like metallaborane products in low or moderate yields.4c,4d The use
here of BH4− instead of BH3 as a reagent with metal halides avoids
cluster expansion reactions. These lower selectivity by giving rise
to a greater variety of metallaborane products.
The reaction of metal halides with borohydrides provides effec-
tive syntheses of metallaborane and hydride complexes. We were
interested to see if a metal pseudohalide would react with BH4− in
a similar way and therefore treated [Mo2Cp2(-SMe)3(-PPh2)] (4)
with NaBH4. Regretably, no reaction (eqn. (7)) was observed under
either the same or more forcing conditions than those employed
for the corresponding reaction of (2) with NaBH4 (section 2.2.2).
Evidently, reactions involving pseudo-halide compounds are less
straightforward than those using halide derivatives.
(10)
Complex 12 had already been obtained (eqn. (6)) by reaction of
−
the bis(chloro) complex 3 with BH4 . Its formation here from the
(7)
tetrakis(nitrile) derivative is understandable. It has been identi-
fied by comparison of its NMR spectra with those of an authentic
sample.
The new derivative 13 was identified as a molybdenum dimer
bridged by both tetrahydroboride and azavinylidene ligands, on
the basis of its spectroscopic data and solid-state structure. 13
combines the structural elements present in complexes 8 (-1 :1-
tetrahydroboride) and 14 (-azavinylidene). Thus, several spectro-
scopic features of 13 are common to 8 or 14. For example, in the
1H NMR spectrum a quadruplet at  7.59 and a doublet at  1.61
2.2.4. Reactions of nitrile complexes 5 and 6 with NaBH4
and of 8 with bases. We have previously shown that the boro-
hydride compound 8 can be effectively synthesised by treating
[Mo2Cp2(-SMe)3(MeCN2)]+ (6) with NaBH4 (eqn. (8)); competi-
tion between this reaction and formation of the azavinylidene 14
can be controlled by the solvent.7 Here we describe the reaction
D a l t o n T r a n s . , 2 0 0 4 , 2 7 0 8 – 2 7 1 9
2 7 1 3

View Article Online
(JH–H = 4.7 Hz) indicate that a hydride has been transferred to one
of the acetonitrile ligands, and that an azavinylidene NC(H)Me
unit is present in the complex. The presence of the BH4 fragment in
13 is also revealed by two typical broad BHb resonances at  −15.57
(intensity 1) and −15.97 (intensity 1), and one broad BHt band at 
1.60 (intensity 2). The results of 2D 1H–1H and 1H–11B correlation
NMR experiments are in accord with these assignments. A 11B-
decoupled 1H–1H 2D-experiment showed that the two resonances at
 −15.57 and −15.97 are coupled to each other and also to the peak
Table 3 Selected bond lengths (in Å) and angles (in °) for [Mo2Cp2(-
SMe)2(-BH4)(-NCHMe)] (13)
Mo1–Mo2
Mo1–S1
Mo1–S2
Mo1–N1
Mo1B1
Mo1–H1×
B1–H1×
B1–H3×
N1–C3
2.630(1)
2.469(1)
2.444(1)
2.083(2)
2.726(2)
1.81(3)
1.25(3)
1.24(3)
1.274(3)
Mo2–S1
Mo2–S2
Mo2–N1
Mo2&B1
Mo2–H3×
B1–H2×
B1–H4×
C3–C4
2.469(1)
2.436(1)
2.094(2)
2.752(2)
1.83(3)
1.12(4)
1.17(3)
1.502(3)
1
at  1.60, and a H–11B inverse-correlation experiment confirmed
1
that the protons are bound to the boron atom. Moreover, the H
Mo1–N1–Mo2
Mo1–S1–Mo2
C3–N1–Mo1
H1×–B1–H2×
H1×–B1–H3×
H1×–B1–H4×
78.1(1)
64.4(1)
138.2(2)
113(2)
109(2)
103(2)
NMR spectrum is typical for a complex with a {Mo2Cp2(-SMe)2}
core; it exhibits two signals at 5.13 and 4.93 ppm for the two cyclo-
pentadienyl ligands indicating a chemically unsymmetrical mole-
cule, and two peaks at 2.07 and 1.02 ppm for the two SMe bridges.
Another interesting spectroscopic feature is that the 11B{1H} NMR
spectrum of 13 shows a broad resonance with a chemical shift
( −27.5) very close to that ( −27.0)7 of the borohydride derivative
8. The structure of 13 (Fig. 7 and Tables 1 and 3) was confirmed
by X-ray analysis of a single crystal obtained from a diethylether
solution at room temperature.
Mo1–S2–Mo2
C3–N1–Mo2
H2×–B1–H3×
H2×–B1–H4×
H3×–B1–H4×
65.2(1)
143.8(2)
109(2)
118(2)
100(2)
is close to that in 13. The azavinylidene ligand spans the Mo–Mo
axis in 13 almost symmetrically [Mo–N 2.083(2) and 2.094(2) Å]
and the N(1)–C(13) distance [1.274(3) Å] accords with bond order
of 2;31 similar features occur in 10, [Mo2Cp2(-SMe)2(-PPh2)(-
NCHMe)] (section 2.2.2), where the Mo–N distance (Table 1,
Fig. 4) is ca 0.03 Å longer than in 13.
2.2. Discussion and mechanistic considerations
The reaction of a metal chloride with a group 13 hydride, e.g.
−
−
BH4 and AlH4 , constitutes one of the standard methods for pre-
paring molecular hydrides.38 Recently, Fehlner et al. proposed that
this reaction proceeds initially by replacement of Cl− by a pseudo-
halide.39 For example, if the first step involves substitution of
−
chloride by BH4 , subsequent loss of the Lewis acid BH3, either by
displacement or hydrolysis, would generate the hydride. The reac-
tions reported here show clearly that this mechanism operates only
in some cases and is strongly dependent on the ancillary ligands.
Thus, Fehlner’s mechanism provides a plausible rationalisation of
the reaction of 2, [Mo2Cp2(-SMe)2)(-Cl)(-PPh2)], with NaBH4
−
in tetrahydrofuran (Scheme 2): the BH4 ligand formed by the
initial replacement of Cl− could generate hydride by loss of BH3
in a second step, initiated by the solvent, which is a Lewis base,
and almost certainly helped by the good donor properties of PPh2.
This last conclusion is supported by consideration of the reaction
of 1, [Mo2Cp2(-SMe)3(-Cl)], with BH4−: the product is the stable
borohydride compound 8. Under similar conditions 2 gives the
hydride complex 9, the presence of PPh2 rather than SMe trans
to the substitution site evidently facilitating displacement of BH3.
Furthermore, no displacement of BH3 is observed, even when 8 is
exposed to CH3CN or bases stronger than tetrahydrofuran, such as
Et3N (see eqn. (9)). Indeed, the -hydrido-tris(-thiolato) analogue
of 9 has never been detected in reactions of 8 with bases, the starting
complex being always fully recovered.
Fig. 7 A view of a molecule of 13.
The molecule contains two CpMo fragments which are almost
symmetrically bridged by two thiolates, a nitrogen atom of the
azavinylidene group and a BH4 unit. There is covalent interaction
between the {Mo2Cp2(-SMe)2(-NCHMe}+ moiety and the BH4
−
anion through two 3c–2e Mo–H–B bonds [Mo–H 1.81(3) and
1.83(3), B–H 1.25(3) and 1.24(3) Å]. There are several structural
parallels with [Mo2Cp2(-SMe)3(-BH4)] (8): thus, the bridging
B–H bonds in 13 are longer than the terminal B–H bonds [1.12(4)
and 1.17(3) Å]; both 13 and 8 contain distorted tetrahedral -1 :1-
BH4 ligands; the Mo–B distances in 13 [2.726(2) and 2.752(2) Å}
are nearly equal and somewhat longer than those in 8 [2.681(6)
and 2.711(6) Å] where direct Mo–B bonding was excluded.7 The
bridging Mo–H bonds in 13 are comparable to those in 8 [1.87(5)
and 1.84(5) Å] but seem somewhat longer than the corresponding
values in [Ir2(C5Me5)2H3(-BH4)] [1.61(8) Å],12 [Mn2(CO)6(-
H)(-BH4)(-Ph2PCH2PPh2)] [1.65(4) and 1.68(4) Å],14a and
[Ru2(C5Me5)2H3(-BH4)] [1.69(3) Å].10c These may indicate that
there is less transfer of hydrogen from boron to the metal atoms in
13 and 8 than there is in the Ir, Mn and Ru complexes, though the
large standard errors on the M–H distances render this conclusion
very tentative. The Mo–Mo distance in 13 [2.630(1) Å] reveals a
bond of unit order (see Table 1).25 It is, however, appreciably longer
than the value of 2.564(1) Å found in the -azavinylidene derivative
[Mo2Cp2(-SMe)3(-NCHMe)] (14),7 a difference which seems to
reflect the replacement of one thiolate group in 14 by a BH4− ligand.
In contrast, the -azavinylidene unit has a similar effect to -thio-
late on the Mo–Mo distance, since the value found in 8 [2.653 Å]7
−
The kinetics of the reaction in which Cl− is replaced by a BH4
when 1 or 2 is treated with NaBH4 depend strongly on the donor
strength of the bridging ligand trans to Cl: thus complete reaction
of 1, with SMe trans to Cl, took 60 h at room temperature, whereas
for 2, where the stronger donor PPh2 is trans to Cl, the reaction was
complete in 1 h under the same conditions. This accords with the
observed structural trans-influence on Mo–Cl bonds: the mean Mo–
Cl(trans to PPh2) distance in 2 [2.536(3) Å] significantly exceeds
the mean Mo–Cl(trans to SMe) bond length [2.483(3) Å] in 1.
It is well established that a metal polyhalide can be transformed
intoapolyborohydridewhichcantheneliminateH2 toyieldametalla-
borane; this reaction in some circumstances competes advantage-
ously with the alternative elimination of a Lewis acid to form a
metal hydride.15 The H2 elimination mechanism has been invoked to
explain the formation of [Cp*2Rh2(-B2H6)], via [Cp*2Rh2(BH4)2],
from Li(BH4) and [Cp*2Rh2Cl4].4e By analogy, we suggest that the
formation of metallaborane [Mo2Cp2(-SMe)2(-B2H6)] (12) by
reaction of the dichloro compound 3 with NaBH4 proceeds through
the bis-(-BH4) derivative C (Scheme 3) although no direct evidence
for the presence of C during the reaction has been obtained.
2 7 1 4
D a l t o n T r a n s . , 2 0 0 4 , 2 7 0 8 – 2 7 1 9

View Article Online
Scheme 2 Proposed pathways for the reaction of compound 2 with
sodium borohydride. [Mo–Mo]=CpMo(-SMe)2MoCp. B=thf.
Scheme 4 Mechanisms of formation of 8 and 14. [Mo–Mo]=CpMo(-
SMe)2MoCp.
studies.41,42 Two possible pathways have been proposed: coordi-
nation of the nitrile to a metal activates the CN group toward
−
hydride addition, either from solvated BH4 (Scheme 5a) or from
a hydrido intermediate which can undergo a 1,3-hydride shift
(Scheme 5b). The latter pathway is a less likely route to 14 since
initial formation of the hydrido compound [Mo2Cp2(-SMe)3(-H)]
cannot be proved. More plausible is the two step process shown
in Scheme 6: initial nucleophilic attack at the CN carbon atom by
H− leads to intermediate D; subsequent transfer of an N lone pair
to the electron-deficient Mo(III) centre and loss of nitrile gives 14.
Scheme 3 Proposed route for the formation of 12. [Mo–Mo]=CpMo(-
SMe)2MoCp.
A more unusual synthetic approach to borohydrides and metal-
laboranes than the treatment of metal halides or hydrides with either
monoboranes or BH4− salts of alkali metals15 involves the substitu-
tion of dative ligands by either BH4− or a polyborane. Only a few
examples of this reaction have been reported:7,10c,40 they include
−
In contrast BH4 and [ReCl2(NCMe)(NO)(PMe3)2] react to give a
−
chelating iminoborane, presumably by BH3 attack on an N lone
pair.42 These results suggest that the molybdenum(III) center in D
(Scheme 6) is a stronger Lewis acid than BH3.
(a) formal substitution of carbonyl by BH4 when Fe(CO)5 reacts
with a mixture of LiBHEt3 and BH3.thf, to give [Fe2(CO)6(B2H6)]40
and (b) formation of borohydride compounds by substituting the
aqua ligands of the dicationic complex [Ru(6-C6Me6)(H2O)3]2+
− 10c
−
by BH4 . More recently, we have shown7 that BH4 reacts with
the bis-nitrile complex [Mo2Cp2(-SMe)3(-MeCN)2]+ to generate
the stable [Mo2Cp2(-SMe)3(-BH4)] compound (8). However,
metal azavinylidene formation competes with formation of 8 and
the tetrahydroboride (8):azavinylidene (14) ratios depend upon
the solvent. Acetonitrile favours the formation of the tetrahydro-
boride (8:14 = 80:20), whereas tetrahydrofuran promotes that of
the azavinylidene product (8:14 = 20:80).7 Here we have shown
that a mixture of 8 and acetonitrile (excess) at room temperature
in tetrahydrofuran does not generate 14 (Scheme 4). Moreover, no
trace of the -hydrido compound [Mo2Cp2(-SMe)3(-H)] has been
detected when 8 is stirred in thf for 24 h. This excludes all possibility
that 14 results from 8 by displacement of BH3 to give a -hydrido
species which subsequently adds acetonitrile (Scheme 4).
Scheme 5 Possible pathways of nitrile reduction.
The reaction of [Mo2Cp2(-SMe)2(-MeCN)4]2+ (5) with NaBH4
also shows the generation of borohydride and metallaborane com-
−
plexes through BH4 or polyborane substitution of dative ligands
(eqn. (10)). Interestingly, both borohydride (13) and polyborane
(12) are probably produced via a polyborohydride intermediate C
and subsequent H2 elimination (Scheme 7). Formation of 12 com-
petes with that of 13. Since 13 combines structural elements present
in 8 (-1 :1-tetrahydroboride) and 14 (-azavinylidene) its forma-
tion logically occurs by consecutive generation of tetrahydroboride
(E) and -azavinylidene (F) intermediates (Scheme 7). The order in
which this happens is unclear and no intermediate has been detected
Instead, we suggest that 14 is formed from 6 by formal addi-
tion of H− onto the C atom of coordinated acetonitrile. Using thf
as solvent favours production of 14.7 Transition metal-assisted
−
reduction of nitriles by BH4 has been the subject of previous
D a l t o n T r a n s . , 2 0 0 4 , 2 7 0 8 – 2 7 1 9
2 7 1 5

View Article Online
Scheme 6 Proposed route for the formation of 14. [Mo–Mo]=CpMo(-
SMe)2MoCp.
1
in the H NMR spectra. Complexes containing both borohydride
and azavinylidene functions resulting from the displacement of two
pairs of dative ligands appear to be without precedent.
3. Conclusions
Some general features of the reactions of thiolato-dimolybdenum(III)
complexes with NaBH4 have been established. Chloro-bridged
compounds give rise to hydroborato or hydride complexes. Abridg-
ing ligand with strong donor properties trans to the leaving chloro
unit favors formation of hydrides. Dichloro-dimolybdenum deriva-
tives similarly give rise to metallaboranes. We have also shown that
hydroborato or metallaborane complexes can be formed if dative
ligands, instead of chlorides, are initially displaced by borohydride
groups. Bis(nitrile) species give rise to either -hydroborato or
-azavinylidene products, while tetrakis(nitrile) derivatives yield
compounds simultaneously containing both of these ligands or,
alternatively, metallaboranes.
4. Experimental
4.1 General procedures and materials
All reactions were performed under an atmosphere of argon or
dinitrogen using conventional Schlenk techniques. Solvents were
deoxygenated and dried by standard methods. Some of the starting
materials [Mo2Cp2(-SMe)3(-Cl)] (1),18 [Mo2Cp2(-SMe)2(-
Cl)(-PPh2)] (2),19 [Mo2Cp2(-SMe)2(MeCN)4] (BF4)2 (5)17 and
[Mo2Cp2(-SMe)3(MeCN)2] (BF4) (6)16 were prepared as described
previously. All other reagents were purchased commercially.
Yields of all products are relative to the starting dimolybdenum
complexes. Column chromatography was carried out with either
silica gel or florisil purchased from SDS. Chemical analyses were
performed either by the Service de Microanalyse I.C.S.N., Gif sur
Yvette (France), or by the Centre de Microanalyses du CNRS,
Vernaison (France). IR spectra were recorded on a Nicolet-Nexus
FT IR spectrometer from KBr pellets. The NMR spectra (1H, 31P,
11B) were recorded at room temperature in CDCl3 or toluene-d8
solutions, unless stated otherwise, with BrukerAC 300 orAMX 400
spectrometers and were referenced to SiMe4 (1H), H3PO4 (31P), and
BF3.Et20 (11B). 1H–1H 2D and 1H–11B 2D experiments were carried
out on a Bruker DRX 500 spectrometer.
Scheme 7 Proposed pathways for the reaction of 5 with NaBH4. [Mo–
Mo]=CpMo(-SMe)2MoCp.
at room temperature. The colour of the solution readily turned from
red to green. The solvent was then removed under vacuum and
the organometallic products were extracted from the residue with
diethylether (2 × 50 ml). After evaporation of the diethylether, the
residue was dissolved in CH2Cl2 (5 ml) and chromatographed on
silica gel. Elution with a mixture of CH2Cl2/hexane (1:1) removed
a green band. Evaporation of the volatiles afforded compound 3 as
an analytically pure green powder (320 mg, 71% yield). This com-
plex 3 was obtained as a mixture of two inseparable anti (3a) and
syn (3b) isomers in. 1:1 ratio. Anal. Calcd. for C12H16Cl2Mo2S2 :C,
29.6; H, 3.3; S, 13.5; Cl, 14.5. Found C, 30.1; H, 3.3; S, 13.6; Cl,
1
14.1%. H NMR /ppm (CDCl3; 298 K), 3a : 5.66 (s,10H,C5H5),
1.60 (s,3H,SMe), 1.54 (s,3H,SMe); 3b : 5.66 (s,10H,C5H5), 1.58
(sS,6H,SMe).
4.2 Preparation of 3
4.3 Preparation of 4 and 7
A solution of complex 517 (700 mg, 0.93 mmol) in CH2Cl2 (50 ml)
was stirred in the presence of 4 equiv. of Et4NCl (616 mg) for 4 h
On standing overnight in ether at 281 K, complex 9 was converted
into compounds 4 and 7. Spectroscopic (1H NMR) analysis of the
2 7 1 6
D a l t o n T r a n s . , 2 0 0 4 , 2 7 0 8 – 2 7 1 9

View Article Online
crop of crystals showed that the two species were each formed in
about 45% yield. 4 and 7 have been characterised by their spectro-
scopic data and the structure of 4 was confirmed by a X-ray diffrac-
tion study. 4:1H NMR /ppm (CDCl3; 298 K):7.3–7.0 (m,10H,Ph),
5.47 (s,10H,C5H5), 1.57 (s,3H,SMe), 1.49 (s,3H,SMe), 0.82
(s,3H,SMe). 31P{1H} NMR /ppm (CDCl3; 298 K):88.7 (s, -PPh2).
7 : H NMR /ppm (CDCl3; 298 K):8.23 (AB,2H,Ph), 7.3–7.0
(m,8H,Ph), 5.57 (s,5H,C5H5), 5.42 (s,5H,C5H5), 2.82 (s,3H,SMe).
31P{1H} NMR (CDCl3;298 K):167.4 (s,-PPh2).
room temperature. The green colour of the solution did not change
after 4 h of stirring. The solvent was then removed under vacuum,
and the organometallic products were extracted from the residue
with diethylether (2 × 30 ml). After evaporation of the diethylether,
the residue was dissolved in CH2Cl2 (5 ml) and chromatographed
on silica gel. Elution with a mixture of CH2Cl2–hexane (1:1) re-
moved a green band. Evaporation of volatiles afforded compound
12 as a green powder (210 mg, 77% yield). 12 was obtained as a
mixture of two inseparable syn (12a) and anti (12b) isomers (in 4:1
ratio) by chromatography. Dark green X-ray quality crystals of 12a
were obtained by cooling a diethylether solution of the complex to
−10 °C for two days. Anal. Calcd. for C12H22B2Mo2S2 :C, 32.5; H,
5.0; B, 4.9. Found C, 32.4; H, 4.9; B, 4.9%. 1H {11B}NMR /ppm
(Tol-d8; 223 K), 12a:4.77 (s,10H,C5H5), 1.84 (s,6H,SMe), 1.11
1
4.4 Reaction of 1 with NaBH4
A solution of compound 1 (500 mg, 1.0 mmol) in tetrahydrofuran
(50 ml) was stirred in the presence of 4 equiv. of NaBH4 (151 mg)
for 60 h at room temperature. The colour of the solution turned from
green to orange. The solvent was then removed under vacuum and
the organometallic derivative was extracted from the residue with
diethylether (2 × 50 ml). Evaporation of the volatiles gave com-
pound 8 as an orange solid (420 mg, 88% yield). 8 was identified by
comparison of its NMR data with those of an authentic sample.7
2
(m,2H,Mo2(-H)4B–Ht), −12.40 (d, JH–H = 16.5 Hz, 4H, Mo–H–
B); 12b : 4.95 (s,10H,C5H5), 1.94 (s,3H,SMe), 1.42 (s,3H,SMe),
2
undetected peaks (2H,B–Ht), −12.02 (d, JH–H = 17.1 Hz, 2H,
2
Mo–H–B), −15.45 (d, JH–H = 17.1 Hz, 2H, Mo–H–B). 11B NMR
/ppm (CDCl3;223 K), 12:44.3. IR (KBr pellet, cm−1), 12:(B–
Ht):2427 m, 2293 f.
Similarly, a solution of compound 4 (250 mg, 0.38 mmol) in
tetrahydrofuran (40 ml) was stirred in the presence of 3 equiv. of
NaBH4 (43 mg) at room temperature. As monitored by H NMR
spectroscopy no reaction was observed, all the starting complex 4
was recovered after 24 hours stirring.
4.5 Reaction of 2 with NaBH4
1
A solution of complex 2 (150 mg, 0.23 mmol) in tetrahydrofuran
(30 ml) was stirred in the presence of 4 equiv. of NaBH4 (35 mg,
0.92 mmol) for 1 h at room temperature. The colour of the solution
readily turned from violet to green. The solvent was then removed
under vacuum and the organometallic product was extracted with
diethylether (2 × 50 ml). Evaporation of the volatiles afforded com-
plex 9 as an analytically pure, green solid (106 mg, 75% yield).Anal.
Calcd. for C24H27Mo2PS2 :C, 47.8; H, 4.5; P, 5.1. Found C, 47.3; H,
4.2; P, 4.9%. 1H NMR /ppm (CDCl3; 298 K):7.3–7.0 (m,10H,Ph),
5.32 (s,10H,C5H5), 1.78 (s,6H,SMe), −8.9 (d, JP–H = 14.4 Hz, 1H,-
H). 31P{1H} NMR /ppm (CDCl3;298 K):154.3 (s,-PPh2).
4.8 Reactions of 8 with Et3N, CH3CN and t-BuNC
Asolution of compound 8 (200 mg, 0.42 mmol) in either tetrahydro-
furan (30 ml) or dichloromethane (30 ml) was stirred in the presence
of 1.5 equiv. of Et3N at room temperature for 14 h. No reaction was
observed and compound 8 was quantitatively recovered.
Similarly, tetrahydrofuran solutions of 8 (0.42 mmol) failed to
react with excess of either CH3CN or t-BuNC in the same conditions
as those employed above for triethylamine.
4.6 Reaction of 9 with CH3CN and t-BuNC
4.9 Reaction of 5 with NaBH4
Compound 9 (50 mg, 0.083 mmol) was dissolved in a mixture
of CH3 CN/thf (1:1) (20 ml). On standing at room temperature
overnight, the colour of the solution turned from green to orange.
Evaporation of the volatiles gave quantitatively compound 10 as
a spectroscopically pure orange powder (53 mg, 0.083 mmol).
Orange X-ray quality crystals of 10 were obtained by cooling a
A solution of complex 5 (400 mg, 0.53 mmol) in acetonitrile
(50 ml) was stirred in the presence of 3 equiv. of NaBH4. (60 mg,
1.59 mmol) for 4 h at room temperature . The colour of the solution
turned from red to maroon. The solvent was then removed under
vacuum and the organometallic products were extracted with dieth-
ylether (2 × 40 ml). Evaporation of the volatiles gave a solid, which
was dissolved in CH2Cl2 (5 ml) and chromatographed on silica gel.
Elution with a mixture CH2Cl2/hexane (1:1) removed a green band.
Evaporation of the volatiles afforded 12a as a green powder (24 mg,
10% yield). Further elution with CH2Cl2/hexane (2.3:1) removed an
orange band, which gave, after evaporation of solvent, compound
13 as an orange powder (76 mg, 30% yield). Further work-up did
not lead to isolation or identification of additional products.
1
concentrated diethylether solution of 10 to −10 °C overnight. H
NMR /ppm (CDCl3; 298 K):8.08 (q, JH–H = 4.6 Hz, 1H, CHCH3),
7.36–7.11 (m,10H,Ph), 5.44 (s,5H,C5H5), 5.28 (s,5H;C5H5), 1.91
(d, JH–H = 4.6 Hz, 3H, CHCH3), 1.24 (s,6H,SMe). 31P {1H} NMR
/ppm (CDCl3; 298 K):76.7 (s,PPh2).
To a solution of complex 9 (70 mg, 0.116 mmol) in tetra-
hydrofuran (10 ml) was added 1 equiv. of t-BuNC (13 l) at room
temperature. The colour of the solution instantaneously turned
from green to orange. The solvent was then removed under vacuum
and the products were extracted from the residue with diethylether
(2 × 20 ml). Evaporation of the volatiles gave a solid, which was
dissolved in CH2Cl2 (3 ml) and chromatographed on florisil. Elution
with CH2Cl2 removed an orange band. Evaporation of the solvent
afforded compound 11 as an orange powder (70 mg, 85% yield).
Anal. Calc. for C29H36Mo2NPS2, 0.25 CH2Cl2 :C, 49.7; H, 5.2; N,
2.0; P, 4.4. Found C, 49.9; H, 5.4, N, 1.9; P, 3.6%. 1H NMR /ppm
(CDCl3; 298 K):11.18 (d, JP–H = 4.4 Hz,1H,CHNt–Bu), 7.10 (d,
JP–H = 5.9 Hz,10H,Ph), 5.30 (s,5H,C5H5), 5.24 (s,5H, C5H5), 1.48
(s,6H,SMe), 1.10 (s,9H,t-Bu). 13C{1H}NMR /ppm (CDCl3;298
K):232.5 (CHNt–Bu), 141.0 (d, JP–C = 36 Hz, Cipso(Ph)), 135.5 (d
JP–C = 7.5 Hz,Ph), 125.1 (s,Cpara(Ph)), 124.8 (d, JP–C = 9 Hz,Ph),
89.3 (Cp), 88.2 (Cp), 53.5 (C(CH3)3), 29.3 (C(CH3)3), 17.5 (d,
JP–C = 18 Hz,SMe). 31P{1H} NMR /ppm (CDCl3; 298 K):92.6
(s,PPh2). IR (KBr pellet, cm−1):(CN):1651 m.
13. Anal. Calcd. for C14H24BMo2NS2 :C,35.5; H, 5.1; N, 2.9.
Found C, 34.3; H, 5.1; N, 2.3%. 1H{11B}NMR /ppm (Tol-
d8;298 K):7.59(q,3JH–H = 4.7Hz, H,NCHCH3), 5.13(s,5H,C5H5),
3
4.93 (s,5H,C5H5), 2.07 (s,3H,SMe), 1.61 (d, JH–H = 4.7 Hz,
3H, NCHCH3), 1.60 (br,2H,BHt), 1.02 (s,3H,SMe), −15.57
(br, 1H, BHb), −15.97 (br, 1H, BHb). 11B NMR /ppm (CDCl3;
298 K):−27.5 (br).
4.10 X-ray crystallographic studies of 1/2, 4, 10, 12a and 13
All crystallographic measurements (Table 4) were made on a Nonius
Kappa CCD diffractometer with Mo–K radiation,  = 0.70173 Å,
typically at 100 K.Absorption corrections were applied. Final least-
squares refinements used all unique F2 with weights chosen to give
a goodness-of-fit near unity.43 Data:parameter ratios range from 18
for 12a to 54 for 13. H atoms bonded to boron were freely refined
for 12a and 13. Otherwise H atoms rode on their parent C atoms.
The quality of the analysis of 1/2 is poor because of the low qual-
ity of the available crystals and the final difference map is noisy in
the vicinity of the metal atoms. Crystals of 12a usually shatter on
4.7 Reactions of 3 and 4 with NaBH4
A solution of complex 3 (300 mg, 0.62 mmol) in tetrahydrofuran
(50 ml) was stirred in the presence of 3 equiv. of NaBH4 (70 mg) at
D a l t o n T r a n s . , 2 0 0 4 , 2 7 0 8 – 2 7 1 9
2 7 1 7

View Article Online
Table 4 Crystal data for complexes 1/2, 4, 10, 12a and 13
1/2
4
10
12a
13
Formula
C37H45Cl2Mo4PS5
1135.6
0.40 × 0.14 × 0.02
Monoclinic
P21/a
C25H29Mo2Mo2PS3
648.51
0.55 × 0.19 × 0.17
Monoclinic
P21/c
C26H30Mo2NPS2
643.48
0.31 × 0.31 × 0.02
Monoclinic
C2/c
C12H22B2Mo2S2
443.92
0.23 × 0.14 × 0.08
Monoclinic
C2/m
C14H24BMo2NS2
473.15
0.4 × 0.3 × 0.05
Orthorhombic
Pbca
M
Crystal size/mm3
System
Space group
a/Å
b/Å
c/Å
15.3646(4)
15.4438(4)
17.1043(5)
98.892(1)
4009.9(2)
4
1.881
100
1.684
9.2543(1)
15.2706(2)
17.6299(3)
98.215(1)
2465.87(6)
4
1.747
100
1.348
16.3590(7)
10.5114(6)
16.3765(7)
116.269(2)
2525.2(2)
4
1.693
120
1.237
11.0564(3)
11.3265(4)
13.7056(4)
109.919(2)
1613.68(9)
4
1.827
293
1.792
9.9585(1)
15.6735(1)
22.1327(2)
/°
/Å3
3454.57(5)
8
1.819
100
1.683
Z
Dc/mg mm−3
T/K
/mm−1
Range /°
I(hkl) collected
Rint
I(hkl) unique/parameters
R1(unique)
wR2(unique)
max, min/e Å−3
2.1–27.5
34406
0.079
8856/447
0.109
0.217
3.5–30.1
22647
0.028
6891/283
0.023
0.051
2.4–30.1
8885
0.055
3653/181
0.068
0.095
3.2–27.5
5835
0.037
1922/102
0.029
0.061
3.4–45.0
39136
0.0262
10699/199
0.0495
0.0804
1.41, −1.07
3.0, −1.5
0.45, −0.59
0.80, −0.81
0.53, −0.37
F. Carrillo-Hermosilla, J. Fernández-Baeza, S. Garcia-Yuste, A. Otero,
A. M. Rodríguez, J. Sánchez-Prada, E. Villaseňor, R. Gelabert, M. Moreno,
J. M. Lluch and A. Lledos, Organometallics, 2000, 19, 3654.
15 T. P. Fehlner, Organometallics, 2000, 19, 2643, and references therein.
16 F. Barrière,Y. LeMest,F. Y. Pétillon,S. Poder-Guillou,P. Schollhammer
and J. Talarmin, J. Chem. Soc., Dalton Trans., 1996, 3967.
17 P. Schollhammer, F. Y. Pétillon, J. Talarmin and K. M. Muir, Inorg.
Chim. Acta, 1999, 284, 107.
cooling and the unit cell volume is doubled at 100 K; the analysis is
therefore based on data measured at room temperature.
CCDC reference numbers 239311–239315.
See http://www.rsc.org/suppdata/dt/b4/b407564a/ for crystallo-
graphic data in CIF or other electronic format.
Acknowledgements
18 M. B. Gomes de Lima, J. E. Guerchais, R. Mercier and F. Y. Pétillon,
Organometallics, 1986, 5, 1952.
We are grateful to N. Kervarec and Dr R. Pichon for the recording
of two-dimensional NMR on a Bruker CRX 500 (500 MHz) spectro-
meter. We thank the CNRS, the EPSRC, Glasgow University and
the University of Brest for financial support.
19 P. Schollhammer, E. Guénin, S. Poder-Guillou, F.Y. Pétillon, J. Talarmin,
K. W. Muir and P. Baguley, J. Organomet. Chem., 1997, 539, 193.
20 (a) N. G. Connelly and L. F. Dahl, J. Am. Chem. Soc., 1970, 92,
7470; (b) W. K. Miller, R. C. Haltiwanger, M. C. Van Derveer and
M. Rakowski DuBois, Inorg. Chem., 1983, 22, 2973; (c) M. McKenna,
L. L. Wright, D. J. Miller, L. Tanner, R. C. Haltiwanger and
M. Rakowski DuBois, J. Am. Chem. Soc., 1983, 105, 5329.
21 (a) P. B. Grebenik, M. L. H. Green, A. Izquierdo, V. S. B. Mtetwa and
K. Prout, J. Chem. Soc., Dalton Trans., 1987, 9; (b) K. Fromm and
E. Hey-Hawkins, Z. Anorg. Allg. Chem, 1993, 619, 261.
22 J. U. Desai, J. C. Gordon, H.-B. Kraatz, V. T. Lee, B. E. Owens-
Waltermire, R. Poli, A. L. Rheingold and C. B. White, Inorg. Chem.,
1994, 33, 3752.
23 J. H. Shin and G. Parkin, Polyhedron, 1994, 13, 1489.
24 H. Adams, N. A. Bailey, A. P. Bisson and M. J. Morris, J. Organomet.
Chem., 1993, 444, C34.
References
1 M. Ephritikhine, Chem. Rev., 1997, 97, 2193.
2 T. J. Marks and J. R. Kolb, Chem. Rev., 1977, 77, 263.
3 X. Lei, M. Shang and T. P. Fehlner, Chem. Eur. J., 2000, 6, 2653.
4 (a) C. Ting and L. Messerle, J. Am. Chem. Soc., 1989, 111, 3449;
(b) Y. Nishihara, K. J. Deck, M. Shang, T. P. Fehlner, B. S. Haggerty
and A. L. Rheingold, Organometallics, 1994, 13, 4510; (c) S. Aldridge,
M. Shang and T. P. Fehlner, J. Am. Chem. Soc., 1998, 120, 2586;
(d) A. S. Weller, M. Shang and T. P. Fehlner, Organometallics, 1999,
18, 53; (e) X. Lei, M. Shang and T. P. Fehlner, J. Am. Chem. Soc., 1999,
121, 1275; (f) M. A. Peldo, M. Shang and T. P. Fehlner, J. Organomet.
Chem., 2000, 609, 77; (g) M. A. Peldo, A. M. Beatty and T. P. Fehlner,
Organometallics, 2003, 22, 3698.
5 (a) H. Brunner, G. Gehart, W. Meier, J. Wachter, B. Wrackmeyer,
B. Nuber and M. L. Ziegler, J. Organomet. Chem., 1992, 436, 313;
(b) S. L. J. Conway, L. H. Doerrer, M. L. H. Green and M. A. Leech,
Organometallics, 2000, 19, 630.
6 J. Ho, K. J. Deck, Y. Nishihara, M. Shang and T. P. Fehlner, J. Am.
Chem. Soc, 1995, 117, 10292.
7 P. Schollhammer, N. Cabon, F. Y. Pétillon, J. Talarmin and K. W. Muir,
Chem. Commun., 2000, 2137.
8 R. Carreño, V. Riera, M. A. Ruiz, C. Bois and Y. Jeannin, Organo-
metallics, 1993, 12, 1946.
9 J. C. Vites, C. Eigenbrot and T. P. Fehlner, J. Am. Chem. Soc., 1984, 106,
4633.
10 (a) L. F. Rhodes, L. M. Venanzi, C. Sorato and A. Albinati, Inorg.
Chem., 1986, 25, 3335; (b) M. A. Esteruelas, M. P. García,A. M. López,
L. A. Oro, N. Ruiz, C. Schlünken, C. Valero and H. Werner, Inorg.
Chem., 1992, 31, 5580; (c) M. Jahncke, G. Meister, G. Rheinwald,
H. Stoeckli-Evans and G. Süss-Fink, Organometallics, 1997, 16, 1137;
(d) D. G. Gusev, F. M. Dolgushin and M. Yu. Antipin, Organometallics,
2000, 19, 3429.
25 F. Y. Pétillon, P. Schollhammer, J. Talarmin and K. W. Muir, Coord.
Chem. Rev., 1998, 178–180, 203.
26 K. W. Muir, P. Schollhammer and F. Y. Pétillon, unpublished results.
27 See for example: (a) A. J. Carty, F. Harstock and N. J. Taylor, Inorg.
Chem., 1982, 21, 1349; (b) C. P. Casey and R. M. Bullock, Organo-
metallics, 1984, 3, 1100; (c) R. P. Rosen, J. B. Hoke, R. R. Whittle,
G. L. Geoffroy, J. P. Hutchinson and J. A. Zubieta, Organometallics,
1984, 3, 846; (d) C. Barré, P. Boudot, M. M. Kubicki and C. Moïse,
Inorg. Chem., 1995, 34, 284; (e) J. E. Davies, N. Feeder, C. A. Gray,
M. J. Mays and A. D. Woods, J. Chem. Soc., Dalton Trans., 2000, 1695;
(f) K. Eichele, R. E. Wasylishen, J. F. Corrigan, N. J. Taylor,A. J. Carty,
K. W. Feindel and G. M. Bernard, J. Am. Chem. Soc., 2002, 124, 1541;
(g) M. E. García, V. Riera, M. A. Ruiz, M. T. Rueda and D. Sáez,
Organometallics, 2002, 21, 5515.
28 A. J. Carty, S. A. Mc Laughlin and D. Nucciarone, Phosphorus-31 NMR
Spectroscopy in Stereochemical Analysis, ed. J. G. Verkade and L. D.
Quin, VCH, Florida, 1987, ch. 16.
29 P. Schollhammer, F. Y. Pétillon, R. Pichon, S. Poder-Guillou, J. Talarmin,
K. W. Muir and S. E. Girwood, J. Organomet. Chem., 1995, 486, 183.
30 F. J. García Alonso, M. García Sanz, V. Riera, A. Anillo Abril,
A. Tiripicchio and F. Ugozzoli, Organometallics, 1992, 11, 801, and
references therein.
11 F. Shafiq and R. Eisenberg, Inorg. Chem., 1993, 32, 3287.
12 T. M. Gilbert, F. J. Hollander and R. G. Bergman, J. Am. Chem. Soc.,
1985, 107, 3508.
13 C. E. Housecroft, Polyhedron, 1987, 6, 1936.
14 (a) R. Carreňo, V. Riera, M. A. Ruiz, Y. Jeannin and M. Philoche-
Levisalles, J. Chem. Soc., Chem. Commun., 1990, 15; (b) A. Antiňolo,
31 P. L. Andreu, J. A. Cabeza, I. del Río, V. Riera and C. Bois, Organo-
metallics, 1996, 15, 3004.
32 (a) A. J. L. Pombeiro and V. Y. Kukushkin, in Comprehensive Coordi-
nation Chemistry II, ed. J. A. Mc Cleverty and T. J. Meyer, Elsevier,
Oxford, 2004, vol. 1, p. 639; (b) K. Fujita, H. Nakaguma, F. Hanasaka
and R. Yamaguchi, Organometallics, 2002, 21, 3749.
2 7 1 8
D a l t o n T r a n s . , 2 0 0 4 , 2 7 0 8 – 2 7 1 9

View Article Online
33 (a) R. D. Adams and N. M. Golembeski, J. Am. Chem. Soc., 1979, 101,
2579; (b) W. J. Evans, J. H. Meadows, W. E. Hunter and J. L. Atwood,
Organometallics, 1983, 2, 1252; (c) D. W. Prest, M. J. Mays and
P. R. Raithby, J. Chem. Soc., Dalton Trans., 1982, 2021.
34 R. M. Siverstein, G. C. Bassler and T. C. Morrill, in Spectrometric
Identification of Organic Compounds, Wiley, New York, 4th edn., 1981,
p. 176.
39 X. Lei, A. K. Bandyopadhyay, M. Shang and T. P. Fehlner, Organo-
metallics, 1999, 18, 2294.
40 G. B. Jacbsen, E. L. Anderson, C. E. Housecroft, F.-E. Hong,
M. L. Bhul, G. J. Long and T. P. Fehlner, Inorg. Chem., 1987, 26,
4040.
41 J. O. Osby, S. W. Heinzman and B. Ganem, J. Am. Chem. Soc., 1986,
108, 67.
35 F. A. Cotton, L. M. Daniels, C. A. Murillo and X. Wang, J. Am. Chem.
Soc., 1996, 118, 4830.
42 A. C. Hillier, T. Fox, H. W. Schmalle and H. Berke, J. Organomet.
Chem., 2003, 669, 14.
36 E. L. Andersen and T. P. Fehlner, J. Am. Chem. Soc., 1978, 100, 4606.
37 R. Poli, Chem. Rev., 1991, 91, 509.
38 J. P. Collman, L. S. Hegedus, J. R. Norton and R. G. Finke, Principles
and Applications of Organotransition Metal Chemistry, University
Science Books, Mill Valley, CA, 1987.
43 G. M. Sheldrick, SHELXL-97, Program for refinement of crystal
structures, University of Göttingen, Germany, 1997; WinGX - A
Windows Program for Crystal Structure Analysis: L. J. Farrugia, J.
Appl. Crystallogr., 1999, 32, 837; DENZO-SCALEPACK: Z. Otwinowski
and W. Minor, Methods Enzymol., 1997, 276, 307–326.
D a l t o n T r a n s . , 2 0 0 4 , 2 7 0 8 – 2 7 1 9
2 7 1 9